In my mind, this is essentially the same issue as Tom and I discussed a short while ago, although it's still not quite clear to me yet. See following link.
http://benefitslink.com/boards/index.php?s...short+plan+yearIf your document is a prototype document, typically, you would amend by restating the adoption agreement by completing the 401k sections and inserting a special 401k effective date. The question is, does the special 401(k) effective date establish a short plan year for the 401k components of the plan or is it simply to define 401(k) eligibility per participant?
The ERISA outline book has several discussions on short plan year issues for 401(k) testing. If there is a short plan year, the administrator may use compensation during the short year OR the administrator may elect to use compensation for the calendar year ending in the plan year per new IRS reg section 1.401(k)-6. If short year compensation is used, then the comp limit is prorated. If there is not a short plan year, then we wouldn't have the opportunity to use only compensation during a short year (unless the document excludes pre-participation compensation). It also explains that a plan may avoid short plan year issues altogether by adopting a retroactive 401(k) effective date back to the beginning of the plan year, even though deferrals do not physically begin until the 401(k) amendment or document is executed.
Of course, the 401(k) component is a disaggregated component. If there is a short plan year, clearly the coverage test would only be run based on census data during the short year and you would disregard any prior terms. This would be the case regardless of whether you elected to use full calendar year compensation. The adoption agreement may allow the plan sponsor to include or exclude pre-participation compensation, but there wouldn't be any pre-participation compensation if there is a short plan year.
If there is not a short plan year, then the question is when the employees are considered eligible. If the effective date of the 401(k) is retroactive back to the beginning of the plan year, then you would include any prior terminated "participants" as though they were benefiting even though they didn't physically have the opportunity to contribute the deferrals. If the special 401k effective date determines everyone's date of entry (but not via short plan year), then this implies all nonexcludable "participants" were nonbenefiting for part of the year, similar to excluding a certain class of employees and then including during the plan year as an eligibile class of participants. Presumably, we would just rely on "daily" coverage testing as opposed to plan year testing to pass coverage (prior to adding the 401k and then on or after adding the 401(k).
You would use full plan year compensation if the plan does not exclude pre-participation compensation. You would use the partial year compensation if pre-participation compensation is excluded. (I believe this is document driven and not elective as it isn't a 414(s) election.) If it is not a short plan year, regardless of whether pre-participation compensation is excluded, you would not prorate the comp limit.
The short plan year treatment issue is still a concern to me because a plan sponsor can avoid prorating the comp limit by simply having a full 12 month "overall" plan year and exclude pre-participation compensation. The question as to whether a special 401(k) effective date establishes a short plan year applies to all 401(k) components of the plan, including match and safe harbor contributions. A short plan year would require us to prorate the comp limit for match or safe harbor nonelective as prototype documents do not typically allow the plan sponsor to specifically elect 12 months compensation for match and safe harbor NE in the event of a short plan year.